Applications for Peptide Arrays

Costs per peptide: From a technical point of view there is one especially prominent feature of our solid-material-based combinatorial synthesis of peptide arrays: We can synthesize very high-density peptide arrays in good quality, with freely chosen sequences, and with posttranslational modifications [5, 6, 8, 9]. Several years ago, peptide arrays cost more than 1 € per peptide ( Currently, such arrays cost <6 cent per peptide – a price that should continuously fall in the forthcoming years concomitantly with a more matured technology ( Especially, the nano3D printer (more) and the one-cavity-one-peptide-method (more) will help to bring further down the cost per peptide in the forthcoming years.

Peptide arrays to find epitopes: The most basic application for cheap peptide arrays is to find the exact epitope on a protein that is bound by an antibody.

Peptide arrays to find antibody fingerprints: However, one peptide could bind different antibody species and one antibody could bind different peptides. If we want to profit from the information from public databases we need to know exactly which amino acid positions within a binding peptide are mandatory for binding – the “antibody fingerprint” (Fig. 8c). Only then do we get a sequence that allows us to query public databases for potential binding partners of this specific antibody.

Characterizing antibodies that reject a transplanted kidney: Fig. 8 shows how we could find out what epitopes / fingerprints on HLA-proteins are targeted in a human that is rejecting a transplanted kidney (collaboration with Prof. Ingeborg Hauser, Univ. Clinic Frankfurt; unpublished).

First, Univ. clinic Frankfurt found out with their standard Luminex assay which variant HLA (= MHC) molecules from the transplanted kidney were targeted by that specific patient serum (DRB1 and DQB1; not shown). Next, we extracted some 3.000 polymorphic 15meric peptides (linear and circular) from the IMGT database that cover all known polymorphic variants from DRB1 and DQB1. These peptides were synthesized with PEPperPRINT’s „peptide laser printer“ and (a) stained with patient serum using secondary anti-human IgG and anti-human-IgM antibodies (seen is one example with anti-human-IgG stained double spots). (b) We then mapped stained peptides to available 3D structures (patient HLA-sequences deviate slightly from this sequence). Surprisingly, all strongly staining peptides fitted to only one region of HLA-DRB1. Even more surprisingly, the same epitope was also targeted by serum from other kidney-rejecting patients. (c) In the next step, we systematically varied found IgG-reactive peptides, and stained again with patient serum to find out which amino acid positions within the peptide are mandatory for IgG antibody binding (the fingerprint; in the example of one specific serum antibody mandatory amino acids within the sequence NSQRDILEDRRGQVD are highlighted in red). (d) Again, we mapped those found amino acids onto the 3D structure of DRB1 to reveal the antibody’s binding epitope.

Fig. 8; Finding a dominant epitope on HLA-molecule DRB1. Both 3D structures are from PDB 5JLZ (Crystal structure of HLA-DRB1*04:01 depicted as 2x dimeric HLA molecules).

Characterizing all of a patient’s antibodies & finding the original protein antigens: Our antibodies shield us from life-threatening infections (but some do harm in autoimmune diseases). If we assume that an amplified antibody needs a concentration of >5µg/ml to protect us from a pathogen, then we can expect some 500 to 2.000 antibody species in a serum that are high enough concentrated to give a signal on a peptide spot. Among these should be the aforementioned protective antibodies.

This basic calculation sparked the idea to characterize all of the antibody species within a patient, in order to learn what antibody fingerprints would be found (Fig. 9): First, we used phage-displayed peptides to identify many different serum antibody binding peptides. However, shortly after publishing these findings, we learned that the used phage displayed peptide library harboured only some 40.000 different peptides. Next, we synthesized identified peptides in the array format [5] and rescreened them with the same serum. Finally, we systematically varied the sequences of validated antibody binding peptides to identify those amino acids within the peptides that are mandatory for binding “their” antibody species. The resulting antibody fingerprints were then used to query data bases to find potential antigens. We did that for one individual and found nearly 50 unique fingerprints that unequivocally represented nearly 50 different antibody species, starting with only 40.000 different random peptides. [13] In comparison, the one-cavity-one-peptide-method (more) will allow us to do such pre-screens with >2,5 Mio random peptides.

Fig. 9; Finding all the antibodies from a patient. In an initial pre-screen peptides displayed on phage were screened for their binding to serum antibodies, immobilized on beads. Next, the identified peptides were re-synthesized as peptide arrays and stained with the same serum. Finally, the validated antibody-binding peptides were systematically varied in order to find the “antibody fingerprints”. These are used to query databases to find those proteins that match the antibody’s specificity.

Interestingly, at least four of found 50 fingerprints were found in an identical form also in the majority of other sera. One of these antibodies obviously binds to the emp protein from Staphylococcus aureus. Currently, we try to find out if this antibody is a kind of natural antibiotic that shields us from this pathogen (collaboration with Ralf Bischoff, dkfz and Dennis Nurjadi, Univ. Clinic Heidelberg).

[5] Stadler V, Felgenhauer T, Beyer M, Fernandez S, Leibe K, Güttler S, Gröning M, Torralba G, Lindenstruth V, Nesterov A, Block I, Pipkorn R, Poustka A, Bischoff FR und Breitling F. (2008) Combinatorial synthesis of peptide arrays with a laser printer Angew. Chem. Int. Ed. 47, 7132–7135; DOI: 10.1002/anie.200801616.

[6] Beyer M, Nesterov A, Block I, König K, Felgenhauer T, Fernandez S, Leibe K, Torralba G, Hausmann M, Trunk U, Lindenstruth V, Bischoff FR, Stadler V und Breitling F. (2007) Combinatorial synthesis of peptide arrays onto a computer chip’s surface. Science 318, 1888; DOI: 10.1126/science.1149751.

[8] Loeffler FF, Foertsch TC, Popov R, Mattes DS, Schlageter M, Sedlmayr M, Ridder B, Dang FX, von Bojničić-Kninski C, Weber LK, Fischer A,Greifenstein J, Bykovskaya V, Buliev I, Bischoff FR, Hahn L, Meier MAR, Bräse S, Powell AK, Balaban TS, Breitling F, Nesterov-Mueller A. High-flexibility combinatorial peptide synthesis with laser-based transfer of monomers in matrix material. Nature Communications, 2016, DOI: 10.1038/NCOMMS11844

[9] Mattes DS, Jung N, Weber L, Bräse S, Breitling F. (2019) Miniaturized and automated synthesis of biomolecules – Overview and perspectives. Advanced Materials, first online published 29 April 2019; 31: 1806656; DOI: 10.1002/adma.201806656

[13] Weber LK, Palermo A, Kugler J, Armant O, Isse A, Rentschler S, Jaenisch T, Hubbuch J, Dubel S, Nesterov-Mueller A, Breitling F, Loeffler FF. (2017) Single amino acid fingerprinting of the human antibody repertoire with high density peptide arrays. Journal of Immunological Methods 443, 45-54; DOI: 10.1016/j.jim.2017.01.012